Regensburg 2022 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
BP: Fachverband Biologische Physik
BP 7: Poster 1
BP 7.40: Poster
Monday, September 5, 2022, 18:00–20:00, P1
Identifying malignant tissue using Laser Induced Breakdown Spectroscopy (LIBS) and Neural Networks — •Elena Ramela Ciobotea1, Christoph Burghard Morscher1, Cristian Sarpe1, Bastian Zielinski1, Hendrike Braun1, Arne Senftleben1, Josef Rüschoff2, and Thomas Baumert1 — 1Kassel Universität, Kassel, Germany — 2Institut für Pathologie Nortdhessen, Kassel, Germany
The problem of differentiating cancerous tissue from a healthy one is currently solved in the diagnostic process through microscopic imaging of stained biopsy sections by pathologists. During surgical removal of cancerous tissue, oncological safety margins must be established to ensure the complete removal of the tumor without affecting much of the neighboring healthy tissue. For this purpose, on-site pathological analysis is done on freshly frozen, stained cuts, which is time consuming. We investigate a new approach to minimize the time of discrimination between malign and benign tissue by an in situ, non-contact spectroscopic analysis. In a proof of principle experiment, a plasma is generated by focusing an 800 nm femtosecond laser on the pathologic postoperative sample. The spectrum of plasma radiation contains information on the element composition of the ablated tissue. Since the recorded spectra are complex and full of information, neural networks are employed to find differences between malign and benign tissue with a high speed and accuracy. This contribution presents the experimental parameters that allow for the best possible differentiation of some biological tissues through fs-LIBS by minimizing deviations between the measurements.